PDA

View Full Version : New U5 mtDNA ancient samples



GailT
02-28-2015, 11:37 PM
Haak et al. published 19 U5 full sequences from ancient remains in their paper "Massive migration from the steppe is a source for Indo-European languages in Europe." I reviewed the U5 mtDNA samples to see which are still found in modern populations and which are extinct. Note that "extinct" in this context means not yet found in modern people, but it is possible that some of these lineages exist but have not yet been discovered. In any case, an "extinct" lineage is one that is extremely rare, if not actually extinct. I also speculated on which subclades of U5 might have an eastern European Hunter Gatherer (EHG) and Yamnaya/Proto-Indo-European origin, and which ones might have had a western Western Hunter Gatherer (WHG) origin.

Lasardis described the Motala samples as Scandinavian Hunter-Gatherer (SHG) and found that the autosomal DNA evidence suggested SHG may not be a source for modern Europeans.

Five of the U5 ancient samples are from the Motala site in Sweden dated at about 7700 ybp. Motala1 and Motala3 are in a proposed new subclade of U5a1 that I'm calling Group L, defined by G5460A. There are 3 modern samples in this group including 1 from Germany and 1 from Finland. Motala1 is ancestral to this group (ignoring 16093) while Motala3 appears to be an extinct lineage of U5a1 Group L. Motala4 and Motala6 represent 2 different extinct lineages U5a2d*, and Motala9 is an extinct lineage of U5a2*.

DEB36 is another extinct U5a1* lineage from the LBK culture in Germany dated at 7200 ybp.

SVP44 from Samarra Russia dated at 7600 ybp is U5a1d* with an extra mutation at 16241 not yet found in modern U5a1d samples,

SVP50 and SVP52 are Yamnaya samples from Samarra Russia dated at 5100 ybp and both are U5a1a1 ancestral.

BZH12 is from Benzigerode-Heimburg, Germany dated at 4200 ybp and is U5a1a2a ancestral.

ESP3 Unetice Culture 3800 ybp is ancestral for U5a1i1a and is common in northern Europe and there are also samples from Hungary and Sri Lanka. U5a1i* is samples from India and Denmark, and U5a1i1* has a sample from Lebanon, so I think it is likely that U5a1i originated in eastern Europe HG and was present among speakers of proto-Indo-European.


ESP5-Corded Ware, Germany is dated at 4400 ybp and is U5a2d* with 6 extra CR mutations. It shares 4 of 6 with a modern sample from Portugal, JQ702144, but ESP5 appears to be extinct.

<b>U5b Ancient Samples</b>

BENZ14 from the Bernburg culture, Germany, dated at 5000 ybp and is ancestral U5a2b4. We have 14 modern samples in this group from Ireland, France, Denmark and Norway. It seems possible that U5a2b4 was WHG in origin.

Mina18a is from Spain dated at about 5750 ybp and is pre-U5b1i* and shares 2 extra mutations with two modern people from Ireland who are also Pre-U5b1i*. Mina18a also has a mutation at 16209 which is unique so it may represent and extinct lineage of this pre-U5b1i* group.

OSH9 is from the Rossen culture in Germany dated at about 6400 ybp and is an extinct lineage of U5b1b* with two unique CR mutations C6644T & A14053G. It's interesting to have another sample of U5b1b* because U5b1b is very widespread geographically and occurs at high frequency in some regions, but it is not very diverse with only two named subclades U5b1b1 and U5b1b2, and a single modern U5b1b* sample from Russia.

QLB6B is from Baalberge, Germany dated at about 5700 ybp and is ancestral U5b2a2. We have 83 modern U5b2a2 samples mostly from Germanic areas of northern Europe (Denmark, UK, Germany, Poland, Scandinavia).

QUEVIII6 is from Germany dated at about 4000 ybp and is pre-U5b2a1b. We have 16 modern U5b2a1b samples mostly from Germanic Europe.


It seems likely that U5a1a, U5a1d and U5a1i originated in EHG who later blended with Neolithc peoples from southwest Asia to form the Yamnaya and who later returned to Europe as part of the migration from the Steppe. The Swedish Motala samples mostly represent extinct mtDNA U5a1 and U5a2 lineages, and the two Motala U5a1* Group L samples are part of a lineage that survives but is rare in Europe today, so this seems consistent with a samll SHG contribution to modern Europeans.

J Man
02-28-2015, 11:40 PM
Would you say that it is probably safe to say now that as a general rule that this is probably the case?

U5a=EHG
U5b=WHG

GailT
03-01-2015, 04:30 AM
Would you say that it is probably safe to say now that as a general rule that this is probably the case?

U5a=EHG
U5b=WHG

No, I don't think so. Some U5a subclades seem to be associated with western Europe, and some U5b subclades seem to have been in eastern Europe at an early date.

Generalissimo
03-01-2015, 05:07 AM
Hey Gail, have you had a look at the non-U Yamnaya mtDNA lineages? Which part of the Near East do they appear to be from? Anatolia, Caucasus, Iran, or some other place?

Krefter
03-01-2015, 05:32 AM
Thanks alot GailT!!!! You're the foremost expert on U5 at Anthrogencia. I've noticed there's a disproportional amount of U5 carriers here.

I'm very surprised that with the small amount of modern U5 FMS samples that some shared private mutations with ancient ones. We can't make many conclusions about SHG, WHG, EHG, etc U5 survival yet because we have so few. Wouldn't you agree that if you picked out 10 random FMS U5s from modern Europe, that it would be hard for them to find matches in your modern data set?

I only did 23andme and HV1+2 at FTDNA and have two private mutations; 215G and 1393A, I could have more. No U5b2a2 at FTDNA has 215G. If I was an ancient person would you say my variant of U5b2a2b1 is extinct? Most people have unfounded private mutations right?

Krefter
03-01-2015, 05:37 AM
The "Ancient mtDNA reference (https://docs.google.com/spreadsheets/d/1HcAhe7QvggT792VruuoZX6IsTg4LhWXV-Z_ZfTk2KGA/edit#gid=1405879894)" has plenty of U5s with private mutations. It'd be interesting to see if what you can find with your modern data base. Do you have a modern database for other haplogroups?

Krefter
03-01-2015, 06:08 AM
The U5b1b* is the most interesting thing to me. Almost all modern ones are U5b1b2 or U5b1b1, yet our only ancient one is neither. There's A 4,000ybp U5b1b from Spain but its study didn't give the sample's mutations. It may have been FMS and also U5b1b*.

Plenty ancient U5bs have 16189C! which is a common mutation and defines U5b1b-h. Although only two have the combined 16189C! and 16192C!, meaning they're likely U5b1b1a-f and are both from Mesolithic Poland. None have 16320T which defines U5b1b1b the only other U5b1b1 clade. U5b1b2 is defined by HV2 mutations so we're out of luck there in terms of ancient mtDNA.

Baltimore1937
03-01-2015, 08:16 AM
I wouldn't be surprised if some ancestral U5b2b2 lineages died out in Norway/Scandinavia, and are now only found in the British Isles (and eventually around the Anglo-speaking world).

Alessio B. Bedini
03-01-2015, 10:12 PM
[..]
OSH9 is from the Rossen culture in Germany dated at about 6400 ybp and is an extinct lineage of U5b1b* with two unique CR mutations C6644T & A14053G. It's interesting to have another sample of U5b1b* because U5b1b is very widespread geographically and occurs at high frequency in some regions, but it is not very diverse with only two named subclades U5b1b1 and U5b1b2, and a single modern U5b1b* sample from Russia.
[..]

I am (with geno 2.0) U5b1b*
I never did, however ,another tests with my mtDNA .
What should I do ?

GailT
03-02-2015, 12:07 AM
I am (with geno 2.0) U5b1b*
I never did, however ,another tests with my mtDNA .
What should I do ?

Try uploading your results to mthap (http://dna.jameslick.com/mthap-new/) first, and see if any extra mutations were found in the Geno 2.0 results.

Krefter
03-02-2015, 06:31 AM
GailT, where can we find your U5 samples with their mutations listed, not just HVR1 and HVR2? How many U5b2a2b1s do you have and where are they from? Do any of them have the CR mutation 1393A? Are there any age estimates for U5b2a2s three subclades?

GailT
05-08-2015, 03:48 AM
GailT, where can we find your U5 samples with their mutations listed, not just HVR1 and HVR2? How many U5b2a2b1s do you have and where are they from? Do any of them have the CR mutation 1393A? Are there any age estimates for U5b2a2s three subclades?


Sorry I missed this question in March - there is one person in the project in U5b2a2b1 with 1393A and at least two more who tested at FTDNA but are not in the project. None of them provide country of origin but I can forward an email to them if you would like to try to get in touch.


For the samples that are in GenBank I used either mitotool or Ian's data for the FMS results. Some people are very concerned about privacy of their CR results so I don't list CR results for project members who are not in GenBank, but I encourage them to submit to GenBank.

GailT
05-08-2015, 03:55 AM
Are there any age estimates for U5b2a2s three subclades?

I'd estimated and age of about 5000 years for each of the 3 subclades, U5b2a2a, U5b2a2b and U5b2a2c. And of course there is the Baalberge ancient sample dated at about 5700 ybp that is ancestral U5b2a2, so that seems generally consistent with the ages of the subclades.

There are 26 U5b2a2b1 FMS samples widely distributed across northern Europe: Germany, Denmark, Scandinavia, Poland, Czech, UK, Wales. Scotland. Germany is the most frequent with 5 samples.

Krefter
05-08-2015, 05:13 AM
I'd estimated and age of about 5000 years for each of the 3 subclades, U5b2a2a, U5b2a2b and U5b2a2c. And of course there is the Baalberge ancient sample dated at about 5700 ybp that is ancestral U5b2a2, so that seems generally consistent with the ages of the subclades.

There are 26 U5b2a2b1 FMS samples widely distributed across northern Europe: Germany, Denmark, Scandinavia, Poland, Czech, UK, Wales. Scotland. Germany is the most frequent with 5 samples.

Thanks for the info! I think just about all those U5b2a2b1s are in my HV1 matches at FTDNA. In my HV1 matches all the FMS people were apart of U5b2a2a or U5b2a2b, and most found a more downstream subclade than that.

It's surprising that our only Neolithic U5b2a2 FMS is negative for all three basal subclades. There is an ~6,000 year old U5b2a2c or pre-U5b2a2c from Neolithic Germany, but that's the most rare subclade.

So, 3 Neolithic U5b2a2 women are the source of the majority of U5b2a2 today, and those three women descended from a distant Upper Palaeolithic or Mesolithic mother from the western half of Europe. I guess most U5b clades in West Europe might follow that narrative.

I would order an FMS test if it wasn't so expensive. But anyways I'm sure FMS U5b2a2s(or of any subclade) will help unravel its recent history. It's exciting to learn some people share private mutations with me. There are certainly many more deep subclades than what's on phylo. I'll be looking at FTDNA, mitosearch and what not to learn more.

GailT
05-13-2015, 02:08 AM
I'm not sure that it would be worthwhile to do the FMS in your case because you might be U5b2a2b1 with 1393A and no additional extra mutations.

GailT
09-10-2015, 04:04 AM
Not sure if there is already a thread on this paper so I'll post the mtDNA results here.
"Ancient genomes link early farmers from Atapuerca in Spain to modern-day Basques"

Eight ancient samples dating from about 5400 to 3600 ybp:
U5a1c1a, U5b1, U5b3, K1a2b, J1c1b1, X2c, H3, H3c

The U5b1 sample is reported as U5b1b but as far as I can tell it is U5b1. None of the samples are U5b1f1a which has been found at very high frequency (12%) among modern Basques. The 5 samples I looked at in more detail (U5 and H3) have extra mutations that seem to be unique, i.e., not yet found in any modern samples, so these might possibly represent lineages that died our. However, the sequencing has some uncertainty as some samples are missing mutations that should be present, and it is not certain that all of these extras mutations are accurate reads,

Baltimore1937
09-10-2015, 04:17 AM
I notice ^ U5b2b is not among them. They must've lived more around Italy back then?.

Krefter
09-10-2015, 04:42 AM
Not sure if there is already a thread on this paper so I'll post the mtDNA results here.
"Ancient genomes link early farmers from Atapuerca in Spain to modern-day Basques"

Eight ancient samples dating from about 5400 to 3600 ybp:
U5a1c1a, U5b1, U5b3, K1a2b, J1c1b1, X2c, H3, H3c

The U5b1 sample is reported as U5b1b but as far as I can tell it is U5b1. None of the samples are U5b1f1a which has been found at very high frequency (12%) among modern Basques. The 5 samples I looked at in more detail (U5 and H3) have extra mutations that seem to be unique, i.e., not yet found in any modern samples, so these might possibly represent lineages that died our. However, the sequencing has some uncertainty as some samples are missing mutations that should be present, and it is not certain that all of these extras mutations are accurate reads,

I have a spreadsheet of Basque mtDNA sequences (https://docs.google.com/spreadsheets/d/1Q1QUJb5d6YgONH6AybM7hnEsHL4Y_WhyrT4jxkWsssY/edit#gid=6408281) from "The Expanded mtDNA Phylogeny of the Franco-Cantabrian Region Upholds the Pre-Neolithic Genetic Substrate of Basques". H3c is the most popular H3 branch in modern Basque and was found in El Potralon. There are a lot of H3*s in Basque to, many of which with extra mutations. We should expect Ancient mtDNA to have extra CR mutations not found in modern samples yet, because that's the case for lots of mtDNA today. Lots of people have private CR mutations that a very small fraction of people share with them. This being the case for an Ancient person doesn't mean their lineage died out or that there's mtDNA discontinuity.


I notice ^ U5b2b is not among them. They must've lived more around Italy back then?.

In total U5b2b probably is only like 1-5% in any given modern population, so we shouldn't expect to find it in a handful of samples. U5b2b can't be defined with HV1 coverage and most ancients are only tested for HVR1. U5b2b has deep subclades that can be defined with HVR1, and there are a few ancients who belong to these subclades.

GailT
09-10-2015, 05:14 AM
We should expect Ancient mtDNA to have extra CR mutations not found in modern samples yet, because that's the case for lots of mtDNA today. Lots of people have private CR mutations that a very small fraction of people share with them. This being the case for an Ancient person doesn't mean their lineage died out or that there's mtDNA discontinuity.

It is certainly possible that these lineages survive but have not yet been found in modern people. There are several U5b1 lineages that are represented by a single modern sample. It would help if we more full sequence samples of modern people, for example, some of the U5 HVR samples reported in the Behar et al. Basque paper.

Krefter
09-10-2015, 05:23 AM
It would help if we more full sequence samples of modern people, for example, some of the U5 HVR samples reported in the Behar et al. Basque paper.

Yeah, HVR doesn't tell much.

Kristiina
09-10-2015, 05:28 AM
It appears that La Brana-Arintero was U5b2c1, so a line close to U5b2b must have been in Iberia already 7000 years ago. http://www.ncbi.nlm.nih.gov/pubmed/22748318

By the way, U5b1 also includes the Saami mtdna line U5b1b1. http://www.ianlogan.co.uk/sequences_by_group/u5b1b1_genbank_sequences.htm

U5b1b: 16144, 16150(?), 16189, 16270 (mentioned by Achilli et.al.(2005). This motif have an estimated age of 8600 ± 2400 years BP and is found both in the Berbers, the Saami and the Yakuts.

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1199377/
http://www.academia.edu/4580559/Complex_Signals_for_Population_Expansions_in_Europ e_and_Beyond

Krefter
09-10-2015, 05:49 AM
It appears that La Brana-Arintero was U5b2c1, so a line close to U5b2b must have been in Iberia already 7000 years ago. http://www.ncbi.nlm.nih.gov/pubmed/22748318

By the way, U5b1 also includes the Saami mtdna line U5b1b1. http://www.ianlogan.co.uk/sequences_by_group/u5b1b1_genbank_sequences.htm

U5b1b: 16144, 16150(?), 16189, 16270 (mentioned by Achilli et.al.(2005). This motif have an estimated age of 8600 ± 2400 years BP and is found both in the Berbers, the Saami and the Yakuts.

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1199377/
http://www.academia.edu/4580559/Complex_Signals_for_Population_Expansions_in_Europ e_and_Beyond

I didn't know that. That's interesting.

U5b1b1a is also found in all of NorthEastern Europe and Scandinavia. I even seen examples from the Middle East.

Baltimore1937
09-10-2015, 08:58 AM
At Ian Logan under U5b2b2 you see 2 with "Bollongino-Neolithic" (I hope I copied that right). Bollongino looks Italian. Is that a geographical location, or the author's name?

http://www.ianlogan.co.uk/sequences_by_group/u5b2b_genbank_sequences.htm

GailT
09-10-2015, 01:59 PM
It appears that La Brana-Arintero was U5b2c1, so a line close to U5b2b must have been in Iberia already 7000 years ago.

Here is what I wrote about U5b2c and the La Brana sample on the U5 project results page (https://www.familytreedna.com/groups/u-5b/about/results) (I need to update this text to include recent FTDNA U5 project samples):


U5b2c has an age estimate of about 15,000 years based on 20 FMS samples. It has been found exclusively in western Europe. There is a one U5b2c* person with ancestry in Ireland.

U5b2c1 has 6 FMS samples including 2 from Spain, and one each from Ireland, England and Germany. One of the Spanish samples is from ancient human remains. Sanchez-Quinto et al. reported a FMS test result for the 7,000 year old remains of a Mesolithic hunter-gatherer at the La Brana-Arintero site which they identified as U5b2c1. Behar et al. estimated U5b2c1 to be about 4000 years old, although with large uncertainty in the date, while my age estimate for U5b2c1 based on the six modern FMS samples is 5,700 years. The La Brana-Arintero sample is at the upper end of the Behar uncertainty range and this raises the question of whether haplogroup ages might be older than estimated by Behar et al., and perhaps the slightly older estimates by Soares et al. might be more accurate. But it is not possible to reach conclusions from a single ancient DNA sample. The presence of U5b2c1 in Ireland and northwest Spain might be indicative of early population exchange between those areas.

U5b2c2 has an age estimate of 4800 years based on 20 FMS samples. This group includes 4 people from Ireland, 2 from Scotland and one each from England and Sweden. It seems likely that U5b2c had its origins in an Iberian or Franco-Cantabrian ice age refuge and arrived in the British Isles at a very early date, based on its frequency and diversity in Ireland.

Jean M
09-10-2015, 02:10 PM
At Ian Logan under U5b2b2 you see 2 with "Bollongino-Neolithic" (I hope I copied that right). Bollongino looks Italian. Is that a geographical location, or the author's name?


It is the author's name. Refers to Bollongino, R. et al. (2013), 2000 years of parallel societies in Stone Age Central Europe, Science, Online October 10 2013.

See http://www.ancestraljourneys.org/europeanneolithicdna.shtml and search for Bollongino.

J Man
09-10-2015, 03:41 PM
Here is what I wrote about U5b2c and the La Brana sample on the U5 project results page (https://www.familytreedna.com/groups/u-5b/about/results) (I need to update this text to include recent FTDNA U5 project samples):

There is also this U5b2c sample from Neolithic Hungary. It probably comes from a European Mesolithic hunter-gatherer ancestress that was assimilated early on into Neolithic farming cultures of the area.

ALP Hungary Polgar-Ferenci-hat [NE1] F 5310-5070 BC

U5b2c 73G, 150T, 263G, 309.1C, 310C, 723G, 750G, 960.1C, 1438G, 1721T, 2706G, 3197C, 4769G, 7028T, 7768G, 8860G, 9477A, 11467G, 11719A, 12308G, 12372A, 13017G, 13617C, 13637G, 14182C, 14766T, 15326G, 16189C, 16270T Gamba 2014

http://www.ancestraljourneys.org/europeanneolithicdna.shtml

J Man
09-10-2015, 03:55 PM
It is the author's name. Refers to Bollongino, R. et al. (2013), 2000 years of parallel societies in Stone Age Central Europe, Science, Online October 10 2013.

See http://www.ancestraljourneys.org/europeanneolithicdna.shtml and search for Bollongino.

Yes..I remember that you had/have an interesting theory that these late ''Fisher-Gatherer'' U5 samples may not actually be locals to the area of Blätterhöhle. They may actually have come from areas further North?

Baltimore1937
09-11-2015, 01:53 AM
My U5b2b2 could be from Angle settlers in England. But don't ask me for specifics; just a guess. They (narrowly related) may be rare or extinct back in Denmark/Schleswig-Hplstein.

Krefter
09-13-2015, 01:02 AM
With FMS samples it's easy to find substructure undefined by phylotree. I added all the U5b2a2 samples Ian Logan has on his site. U5b2a2 spreadsheet (https://drive.google.com/open?id=18W7IMqKtQSVAUySDZfy71ScqPsmQvohefei1mWQ8D aQ). GaiT would you mind at some point sharing FTDNA U5 data in the future?

I found sub-structure in U5b2a2 not shown on phylotree. Each is represented by at least 2 samples. Each mutation in each lineage might be 1,000s of years apart and more samples can define the clades better. Most U5b2a2b is U5b2a2b1 and U5b2a2b2(-709A). The Polish U5b2a2d from Ian Logan has three HVR1 matches in my data from Poland and BeloRussia. One U5b2a2b* has a single extra mutations 16172C, the same extra mutation an over 10,000 year old Mesolithic U5b2a2 from Germany had. It could mean U5b2a2b is over 10,000 years old.

It looks like most popular form of U5b2a2 is U5b2a2b or specifically U5b2a2b1. In my data U5b2a2 appears to be pretty much non-existent everwhere except North/Central/East Europe. GailT says the same about his FTDNA data. Looking at FMS from FTDNA and Ian Logan might reveal some interesting substructure in U5b2a2.

I'm definitely going to use Ian Logan's data a lot, it's very useful. I'm want as much full-sequence data as possible.

U5b2a2d: T16217C C16234T 16192!
U5b2a2b2: 709A
U5b2a2b2a: C3510T
U5b2a2b2b: 9305A 9758C

U5b2a2b1a: T16192C! C16193T
U5b2a2b1b: A13722G
U5b2a2b1c: 1393A

Emmish
11-06-2019, 03:40 PM
Hi! I am from Sweden. My mom has tested with FamilyTreeDNA and her haplogroup is U5b2a2b.

There isn't very much to read about it online, but what I have read here is very useful. I would like some advice on how I find out which path my maternal ancestors to reach Sweden. I find the tools in FamilyTree to be a bit unclear. It just shows the main haplogroups on a map with the years they might have arisen. I used the mthap tool, but the result still didn't give me anything before haplogroup R.

There are so many tools for autosomal tests but there doesn't seem to be that many for mtDNA FASTA files. Tips and advice on how to find out more would be appreciated. Thanks!

solarius
11-15-2019, 04:27 AM
Hi! I am from Sweden. My mom has tested with FamilyTreeDNA and her haplogroup is U5b2a2b.

There isn't very much to read about it online, but what I have read here is very useful. I would like some advice on how I find out which path my maternal ancestors to reach Sweden. I find the tools in FamilyTree to be a bit unclear. It just shows the main haplogroups on a map with the years they might have arisen. I used the mthap tool, but the result still didn't give me anything before haplogroup R.

There are so many tools for autosomal tests but there doesn't seem to be that many for mtDNA FASTA files. Tips and advice on how to find out more would be appreciated. Thanks!

https://yfull.com/mtree/U5b2a2b/

[UPDATED with regards to submitting samples to Genbank on behalf of relatives]

MtDNA haplogroups are incredibly old due to how slowly mitochondrial DNA mutates, did you browse FTDNA's learning centre? It is actually quite useful for laymen in my opinion.

Did you do a full mitochondrial genome sequencing test?

Would you be interested in donating your data for scientific research? The US based National Center for Biotechnology Information is happy to accept and anonymously publish your sample data for other scientists worldwide to use:

https://www.ncbi.nlm.nih.gov/genbank/

Such samples are used to define the haplotree and identify mutations for each group, to which your sample can help with, below is one such paper that used publicly available Genbank samples to help define the tree:

https://www.ncbi.nlm.nih.gov/pubmed/22482806

A renown amateur geneticist and retired physician Ian Logan has a website that helps you make such submissions with a tool (you will need your mother's consent and for her to send it from her email if you wish to do this, NCBI will remove/reject samples submitted on behalf of relatives):

http://www.ianlogan.co.uk/checker/submission_maker.htm

If you do think that you will make such an important donation, do inform him once you obtain an ascension number from NCBI's Genbank, as he maintains a mitochondrial haplotree on his website, including sample groupings:

http://www.ianlogan.co.uk/sequences_by_group/u5b2a2_genbank_sequences.htm